[PETERSON] THE ROMANCE OF A MANUSCRIPT 485 
his day. He belonged to Liége, and it seems not improbable that 
Liége was the first resting place of the Cluni manuscript after it had 
started out on its travels. Its later history may be connected with 
the three letters of the alphabet, N. F. and M., as they are used in the 
great critical edition of Cicero’s Speeches by Baiter and Halm. 
N stands for Nannius, which is the Latin name for Pierre Nanning 
of Louvain; ob. 1557. Nanning reports various readings from a 
Codex which he calls ‘aureus libellus’,—a golden book, and these various 
readings were at once accepted by contemporary editors, and were 
finally incorporated in the text of Lambinus. I can now state with 
certainty that the Cluni Codex is the manuscript from which these 
variants were taken. 
F stands for Fabricius which is the Latin name for Schmidt (of 
Diiren; ob. 1573). This Schmidt supplied Lambinus with variants 
for his epoch-making edition, and these were added after Lambinus’s 
death, in the margin of the second issue of that work, always with 
the reference, —V.C. (=Vetus Codex). The source of these various 
readings can now be stated to have been the Cluni Codex. 
M stands for Metellianus, a Codex owned by Jean Matal (1520- 
1597) and used also by another collator called Gulielmius. To each 
of these the text of Cicero is under great obligation, and it was for a 
long time contended that they must have used different manuscripts. 
It can now be shown, however, that both used the same manuscript, 
and that this was the Cluni Codex. 
The formula which may now be adopted with confidence is 
accordingly: 
INGE Vie 3G 
The Cluni Manuscript is, however, now in a very mutilated 
condition. It is in fact only a shadow of its former self,—a thin and 
greatly emaciated volume instead of a lordly and bulky codex, like 
its sister of Paris. But here again good luck followed the effort to 
restore it by the help of its progeny. The portion which is most 
defective is that which formerly contained a complete version of two 
of the books into which the Speeches against Verres are divided. 
Scholars had already noted that a certain manuscript of the 15th 
century, now at Florence (Lg. 42), while setting forth for the rest of 
these speeches a text by no means above the average, contained for 
these particular books a tradition that was easily recognized as some- 
thing older and more free from corruption. This part of the manu- 
script in question can now be shown to have been copied directly from 
the Cluni Codex while it was still intact, so that although the latter 
is now merely a fragment, we are able to restore the missing parts 
from this transcript, and can readily understand now why the tradition 
